Vibration Monitoring of Turbine Blade Using Image Processing

概要

論文の詳細を見る
Among vibrations to occur in a turbine, vibrations of a main shaft or a blade are very dangerous. A lot of systems monitoring vibrations to occur to a main shaft were developed, and they have been commercialized. On the other hand, it is very difficult to measure vibration of a turbine blade during operation because blades are united with a main shaft and turn. In this study, we proposed the technique to measure vibration of a turbine blade in a turn using an image processing. By using the CCD camera, we captured the end face of a blade in a turn and determined a position of a blade from an image data, and then estimated the amplitude of vibration statistically. The movement of the blade at high speed made a captured image blur, and degraded the estimation of the position of the blade. By processing the blurred image as the inverse problem, we could estimate the position of a blade precisely. One piece of image data was provided by a turn of once. Because the natural frequency of a blade was higher than the rotational speed, we were requested to plot the probability density of the position of the blade for estimating amplitude of the vibration. Under a supposition that the blade vibrates sinusoidally, the probability density for the blade vibration can be calculated. By comparing the probability density for sine wave with the processed probability density for the blade, the amplitude of blade vibration could be evaluated. The rotating speed of the blade and the shape of an end face of a blade affected the estimation of the position of the blade. We checked these effects and verified the high accuracy of our technique experimentally.

Vibration Monitoring of Turbine Blade Using Image Processing

スポンサーリンク

概要

著者

関連論文

スポンサーリンク